"I googled that for you" is generally deployed when you know better than the other person (and likely wish to rub it in their face that they should know better too).
I feel like the LLM equivalent here sort of demonstrates the exact opposite (I don't know enough about this topic to even doubt the accuracy of the machine...)
I often get to a point where it's faster to tell a person to just look at the output of the ai agent than to write it out myself.
"I googled that for you" can also be done from a position of ignorance too.
This is just a new thing that new cultural norms are developing from.